Nonlinear Geophysics [NG]

NG42B
 MC:2014  Thursday  1142h

Scale, Scaling, and Nonlinear Variability in Space-Time Precipitation: Data, Measurements, Models, and Theories II


Presiding:  D Schertzer, ENPC; A Montanari, Univeristy of Bologna

NG42B-01 INVITED

Spatial Downscaling of Rainfall: Does it Matter for Flood Forecasting?

* Krajewski, W F witold-krajewski@uiowa.edu, The University of Iowa, IIHR-Hydroscience & Engineering, Iowa City, IA 52240, United States
Mandapaka, P V pradeep-mandapaka@uiowa.edu, The University of Iowa, IIHR-Hydroscience & Engineering, Iowa City, IA 52240, United States
Mantilla, R ricardo-mantilla@uiowa.edu, The University of Iowa, IIHR-Hydroscience & Engineering, Iowa City, IA 52240, United States

Many downscaling techniques have been proposed to address the mismatch between the spatial and temporal scales of rainfall by some observational systems or numerical simulations, and the requirements of land surface hydrologic models. Most of these techniques focus on the mechanism producing the great variability in both space and time that is characteristic of rainfall. In this study the authors investigate the hydrologic consequences of the downscaling. The hydrologic scope of the study is limited to flooding and thus focuses on the scale dependence of peak flows with respect to drainage area. The authors use numerical simulations using a rainfall-runoff model based on a landscape decomposition framework that preserves the major elements of flood genesis, i.e. runoff generation at hillslopes and channel routing of the runoff. The decomposition preserves the full extent of the channel drainage network and its scaling properties. The framework allows addressing the questions of appropriate scale at which the uncertainty associated with the downscaling is sufficiently reduced. The framework also offers an insight into the role of the landscape in partitioning rainfall and the associated mass, momentum, and energy.

NG42B-02 INVITED

Multifractal Rainfall: Utopia or Reality?

* Tchiguirinskaia, I Ioulia.Tchiguirinskaia@cereve.enpc.fr, CEMAGREF, 3275 route de Cezanne, Aix-en-Provence, 13182 Cx5, France
* Tchiguirinskaia, I Ioulia.Tchiguirinskaia@cereve.enpc.fr, Universite Paris-Est Ecole des Ponts ParisTech CEREVE, 6-8 Avenue Blaise Pascal Cité Descartes, Marne-la-Vallee, 77455 Cx2, France
Schertzer, D J Daniel.Schertzer@enpc.fr, CNRM/GAME Meteo-France & CNRS, 42 Av, G. Coriolis, Toulouse, 31057 Cx, France
Schertzer, D J Daniel.Schertzer@enpc.fr, Universite Paris-Est Ecole des Ponts ParisTech CEREVE, 6-8 Avenue Blaise Pascal Cité Descartes, Marne-la-Vallee, 77455 Cx2, France
El Tabach, E el-tabae@cereve.enpc.fr, Universite Paris-Est Ecole des Ponts ParisTech CEREVE, 6-8 Avenue Blaise Pascal Cité Descartes, Marne-la-Vallee, 77455 Cx2, France
Lovejoy, S lovejoy@physics.mcgill.ca, Phys. Dept. U. McGill, 3600 University St., Montreal, QUE H3A 2T8, Canada

Throughout the world, the development of multifractals brought out brand new techniques to handle extreme variability over a wide range of space-time scales. The resulting stochastic simulations are defined with the help of a very limited number of parameters and are able to reproduce this complex behavior, in particular long range dependencies, the clustering of extremes and fat tailed probability distributions. However, the relevance of multifractals are still often questioned in the name of claimed large uncertainties on the multifractal paramater estimates and lack of objective test of scaling behavior. We therefore investigate the ability of using relatively short or incomplete rainfall data records to obtain reliable statistical predictions and to evaluate their uncertainty. We consider three main aspects of this evaluation: the scaling hypothesis adequacy, the multifractal parameter estimation error and the quantile estimation error. We first use multiplicative cascade models to generate long series of multifractal data. The parameters of the simulated samples are chosen to cover the dispersion of the rainfall multifractal parameter estimates in the scientific literature. Splitting these long multifractal series into shorter and shorter sub-samples, we defined a metric for parameter estimation error. The estimated parameters define semi-analytical quantiles for a range of excedance probabilities. The distribution of these quantiles with respect to the theoretical ones yield estimates of the quantile estimation error. This can be used to obtain confidence intervals for multifractal predictions. The obtained metrics were applied to several rainfall time series from a Paris suburb area. Using GIS and a physically based and spatially distributed numerical model of surface runoff and subsurface flows, we illustrate potential consequences of non-adequate rainfall predictions in urban hydrology, as well the relevance of multifractal rainfall modelling.

NG42B-03

Climate Change and Hydrological Extreme Evolution Through a Multifractal Analysis of a Mesocale Model

* Gires, A Auguste.Gires@cereve.enpc.fr, Universite Paris-Est Ecole des Ponts ParisTech CEREVE, 6-8 Avenue Blaise Pascal Cité Descartes, Marne-la-Vallee, 77455 Cx2, France
Schertzer, D J Daniel.Schertzer@enpc.fr, CNRM/GAME Meteo-France & CNRS, 42 Av. G. Coriolis, Toulouse, 31057 Cx, France
Schertzer, D J Daniel.Schertzer@enpc.fr, Universite Paris-Est Ecole des Ponts ParisTech CEREVE, 6-8 Avenue Blaise Pascal Cité Descartes, Marne-la-Vallee, 77455 Cx2, France
Tchiguirinskaia, I Ioulia.Tchiguirinskaia@cereve.enpc.fr, CEMAGREF, 3275 route de Cezanne, Aix-en-Provence, 13182 Cx, France
Tchiguirinskaia, I Ioulia.Tchiguirinskaia@cereve.enpc.fr, Universite Paris-Est Ecole des Ponts ParisTech CEREVE, 6-8 Avenue Blaise Pascal Cité Descartes, Marne-la-Vallee, 77455 Cx2, France
Royer, J jean-francois.royer@meteo.fr, CNRM/GAME Meteo-France & CNRS, 42 Av. G. Coriolis, Toulouse, 31057 Cx, France
Dufresne, J Jean-Louis.Dufresne@lmd.jussieu.fr, Laobratoire de Meteorlogie Dynamique CNRS & UPMC, U. P.& M. Curie Case 99 4 Place Jussieu, Paris, 75252 Cx05, France
Desplat, J julien.desplat@meteo.fr, DIRIC/BE Meteo-France, 2 Av. Rapp, Paris, 75007, France
Lovejoy, S lovejoy@physics.mcgill.ca, Phys. dept. McGill U., 3600 University St., Montreal, QUE H3A 2T8, Canada

In its last report, the IPCC (http://www.ipcc.ch) emphasizes the question of scales and the necessity to obtain much finer resolutions for hydrological processes in climate scenarios to assess the time evolution of the hydrological extremes. One may bridge up the gap between meteorological and hydrological scales with the help of downscaling techniques, which are statically or/and physically based. In particular, one may exploit the scaling properties of the precipitation to downscale it either numerically by stochastic subgrid modeling or theoretically with the help of a few scaling exponents (Royer et al. C.R. Geoscience, 340, 2008). However, we first suggest that these techniques need to be validated not only with the help of empirical data of the actual climatology, but also by comparison with mesoscale numerical simulations in normal and pertubated conditions. Secondly, we present how to proceed to such a comparison with the help of a multifractal analysis of the hydrometeorological fields of the Meso-NH model (Meteo-France/CNRM and Laboratoire d'Aérologie, Toulouse, France), a model which has been rather extensively used for mesoscale research and should shortly become operational.